Customized Spending - Proper Assignment of Output

Hello Andrew. We are trying to clearly understand your project and questions. What are your categories of spending? Are you dividing your students into undergraduate and graduate students? Because graduate students will have different spending patterns than under graduate students. Are you further dividing your student population in to resident and non-resident students? Are you removing student spending from university operations? Because in most universities, student spending (tuition and board) are included in university operations. If you don’t remove from operations, then you will be double counting the impact of students in your project. These are just several questions to consider. Thus, rather than providing you with information that may not address your question, we would appreciate if you could help us understand a little bit more about what you are trying to accomplish. Likewise how detailed is your data? Generally speaking you would like to tie the information down to specific sectors if possible, which it sounds like you have. On the surface, option #2 seems to be more appropriate because you can edit the Institution Spending Pattern to include your spending and remove those expenditures that are not relevant to college spending. Finally, is the source of your data the University’s office of Institutional Research? If so, that is a good place to start. Any additional information you can provide will help us in providing better support to you. Thanks!

Hello, Thank you for your response and I apologize for not providing those details earlier. We are using cost of attendance data from the University's Office of Financial Aid and Office of Institutional Research. The data consist of room & board (only for off-campus students), transportation, books, and other miscellaneous personal expenses for both undergraduate and graduate expenses for FY2013. Since we are separately analyzing the operating expenditures, I have made sure to remove all R&B expenses for on-campus students so as to not double-count those expenses in the operating analysis. Also, since we are using the IMPLAN data set for only one state, I have ensured that all transactions that took place out of the state (including operating, construction and payroll for out-of-state employees) have been removed from the analysis. Our student expenditure data are not too detailed but they consist of the total estimates of student spending in those categories above broken down for undergraduate and graduate students on each campus. I have already selected commodity codes for the sectors of student spending and will use state appropriations and state funding as direct inputs to the model. I am now trying to find out how to appropriately assign the outputs of student spending to their proper indirect and induced categories. I expect that it is not as simple as just adding up those effects and lumping them into the induced category, so I am wondering if there is a specific formula or recommendation that you have for manually calculating this type of effect. Thanks again! Andrew

Hello Andrew For the non-room and board student expenditures, we would recommend doing a Household Spending Pattern using the lowest income category, edited to remove 361, 360 (if the students live on-campus), and any other purchases you don't think students would make (again for on campus students you might also remove utilities). Be sure to re-normalize the spending pattern after any edits so that the Sum of Event Values is 1.00. (Events Options> Change All> Normalize Events) The reason for this approach is that even your known expenditure categories are a little broad. For instance, "Transportation" could be public transportation, taxis, gasoline for their own cars, etc. Likewise for "Books" - did they get them at a used book store, new book store, or on-line? Each of these is a different Sector in IMPLAN. Rather than you deciding how to split them up, you could just let IMPLAN split them up based on the average low-income household's spending behavior. Regarding the fact that this method results in Direct, Indirect, and Induced impacts, this is okay and as it should be. These purchases are Final Demands, as opposed to purchases that are associated with being employed by a certain industry, and thus should have Direct, Indirect, Induced impacts. The Household Spending Patterns are designed that way for exactly this type of use - when new households come in (say tourists, population growth, retirees, students) who spend their money there but that money is not associated with employment in the area - i.e., it is income and spending that is not "induced" by an Industry Change Activity. Please let us know if you have any additional questions.